UNIVERSITY OF SYDNEY
FACULTY OF LAW
EXAMINATION FOR THE DEGREE OF MASTER OF CRIMINOLOGY
MASTER OF LAWS, MASTER OF JURISPRUDENCE
S.J.D.
CRIME, RESEARCH AND POLICY 1
JUNE 1996
TIME ALLOWED: THREE HOURS AND FIFTEEN
MINUTES (INCLUDING READING TIME)
OPEN BOOK
CANDIDATES MAY BRING IN A HAND HELD CALCULATOR BUT MUST SHOW ALL CALCULATIONS
FORMULAE AND STATISTICAL TABLES WILL BE PROVIDED
CANDIDATES SHOULD ANSWER TEN (10) QUESTIONS IN TOTAL -
TWO (2) QUESTIONS IN PART A, ALL QUESTIONS IN PART B AND THREE (3) QUESTIONS IN PART C
PART A (20 marks)
Write short essays on two (2) questions in this section. Each question is worth 10 marks.
QUESTION 1
You are commissioned by the Minister for Aboriginal Affairs to determine the level of literacy among Aboriginal inmates in NSW prisons as compared with non-Aboriginal inmates. How would you go about designing this research? What factors would be important in determining the methodology you would use?
QUESTION 2
Global Capital Inc. have been operating a private prison in NSW for five years. In tendering for a new private prison to be built in Victoria they present a number of arguments stressing that their private prison in NSW is superior to other non-private prisons in that state. They claim that their private prison has a lower recidivism rate than other NSW prisons. As an analyst for the Victorian government you are required to assess the research on which this claim is based. What factors would you take into account in assessing the research? What are some of the likely problems which would face a researcher who wanted to compare recidivism rates for two different prisons?
QUESTION 3
The Minister of Education is interested in the nexus between school truancy and juvenile offending. You are given the task of designing a research program to consider this issue. One aspect of the research program which you devise is an ethnographic study of school truants. Your supervisor is not familiar with ethnographic research. Provide a justification of your decision to undertake that aspect of the research.
QUESTION 4
The Solomon four group design offers a number of advantages in terms of experimental design but in reality is rarely used in criminological research. What are the strengths and the limitations of this design?
PART B (65 marks)
All questions in this section are compulsory
QUESTION 1 (10 marks)
As a defence lawyer preparing an appeal you calculate that your client's sentence may be represented as a z score of 2.6.
a) What is the value of standardising your client's score as a z score?
b) What descriptive statistics does the z score combine?
c) What percentage of offenders would have been given a sentence longer than your client?
QUESTION 2 (15 marks)
Policing in NSW is administered in four distinct regions. The NSW Police Service is concerned that policing practices differ across the four police regions to an extent which is not justified. They ask a researcher to compare clear up rates for several offences categories for the four regions. The following data concerning common assaults are collected for December 1995 for each of the four police regions:
South South West North North West Total
Region Region Region Region
Common
Assaults
Cleared 181 288 341 390 1200
Not Cleared 219 162 284 135 800
Total 400 450 625 525 2000
a) Which is the dependent variable and which is the independent variable in this research?
b) What test would you use to determine if there is a relationship between the two variables? Why?
c) Calculate the test named in b).
d) What interpretation would you give to the result calculated in c) for an alpha level of 0.05?
QUESTION 3 (5 marks)
A tabloid daily newspaper "The Exaggerator" includes the following news item:
"A random sample survey of local residents indicates that 15,300 people were victims of assault in the last year."
The newspaper's use of the data indicates a failure to grasp some important issues in sample research. In what way is the newspaper's presentation of the data misleading? How could it be improved?
QUESTION 4 (10 marks)
The Judicial Commission in NSW is interested in determining the attitudes of judges and magistrates to judicial education.
a) How large a sample would they need to determine this question at an alpha level of 0.01, to within 3%. Pilot work in Victoria suggests that 32% of judges and magistrates favour judicial education.
b) The sample survey described above is undertaken by the Judicial Commission, using the sample size calculated in a). They find that in NSW 28% of magistrates and judges sampled are in favour of judicial education. Within what range does the population value lie (alpha = 0.01)?
QUESTION 5 (20 marks)
A researcher is interested in the relationship, if any, between the length of time that a heroin user has been regularly using heroin and the number of convictions that person has. The researcher collects the following data:
Individual Length of drug use Number of convictions
1 2 5
2 8 9
3 9 12
4 5 7
5 7 4
6 3 2
7 12 13
8 10 8
9 8 10
10 6 5
a) What test would you use to determine whether there is a relationship between the two variables? Why?
b) Calculate the test named in a)
c) What is the coefficient of determination? What does it mean in this example?
d) Would you use a one tailed test or a two tailed test? Why?
e) With an alpha level of 0.05, and your answer in d) how would you interpret this result?
PART C (15 marks)
YOU ARE REQUIRED TO ANSWER THREE (3) QUESTIONS IN THIS SECTION. EACH QUESTION IS WORTH FIVE (5) MARKS.
QUESTION 1
In examining the relationship between two interval level variables you calculate a Pearson's correlation of 0.69. What additional information do you need before you can go on to interpret what the correlation means?
QUESTION 2
What is a Type II error, and how can you reduce its probability?
QUESTION 3
What is a two tailed test?
QUESTION 4
How do you measure the relationship between two variables which are each measured using different scales of measurement?
QUESTION 5
What is the null hypothesis and why is it important?
QUESTION 6
Can you use a Spearman's rank order correlation for interval level data? Why or why not?
QUESTION 7
What factors are important considerations in determining the size of a random sample needed to undertake a given piece of research? Why ?
QUESTION 8
What does a negative z score indicate?
QUESTION 9
Why should we examine whether interval data is skewed before determining which measure of central tendency to use?
UNIVERSITY OF SYDNEY
FACULTY OF LAW
EXAMINATION FOR THE GRADUATE DIPLOMA IN CRIMINOLOGY
CRIME, RESEARCH AND POLICY 1
JUNE 1996
TIME ALLOWED: THREE HOURS AND FIFTEEN
MINUTES (INCLUDING READING TIME)
OPEN BOOK
CANDIDATES MAY BRING IN A HAND HELD CALCULATOR BUT MUST SHOW ALL CALCULATIONS
FORMULAE AND STATISTICAL TABLES WILL BE PROVIDED
CANDIDATES SHOULD ANSWER NINE (9) QUESTIONS IN TOTAL -
TWO (2) QUESTIONS IN PART A, ALL QUESTIONS IN PART B AND THREE (3) QUESTIONS IN PART C
PART A (30 marks)
Write short essays on two (2) questions in this section. Each question is worth 15 marks.
QUESTION 1
You are commissioned by the Minister for Aboriginal Affairs to determine the level of literacy among Aboriginal inmates in NSW prisons as compared with non-Aboriginal inmates. How would you go about designing this research? What factors would be important in determining the methodology you would use?
QUESTION 2
Global Capital Inc. have been operating a private prison in NSW for five years. In tendering for a new private prison to be built in Victoria they present a number of arguments stressing that their private prison in NSW is superior to other non-private prisons in that state. They claim that their private prison has a lower recidivism rate than other NSW prisons. As an analyst for the Victorian government you are required to assess the research on which this claim is based. What factors would you take into account in assessing the research? What are some of the likely problems which would face a researcher who wanted to compare recidivism rates for two different prisons?
QUESTION 3
The Minister of Education is interested in the nexus between school truancy and juvenile offending. You are given the task of designing a research program to consider this issue. One aspect of the research program which you devise is an ethnographic study of school truants. Your supervisor is not familiar with ethnographic research. Provide a justification of your decision to undertake that aspect of the research.
QUESTION 4
What are the features of experimental research? What are some of the difficulties of undertaking experimental research in the area of criminal justice?
PART B (55 marks)
All questions in this section are compulsory
QUESTION 1 (10 marks)
As a defence lawyer preparing an appeal you calculate that your client's sentence may be represented as a z score of 2.6.
a) What is the value of standardising your client's score as a z score?
b) What percentage of offenders would have been given a sentence longer than your client?
QUESTION 2 (15 marks)
Policing in NSW is administered in four distinct regions. The NSW Police Service is concerned that policing practices differ across the four police regions to an extent which is not justified. They ask a researcher to compare clear up rates for several offences categories for the four regions. The following data concerning common assaults are collected for December 1995 for each of the four police regions:
South South West North North West Total
Region Region Region Region
Common
Assaults
Cleared 181 288 341 390 1200
Not Cleared 219 162 284 135 800
Total 400 450 625 525 2000
a) Which is the dependent variable and which is the independent variable in this research?
b) What test would you use to determine if there is a relationship between the two variables? Why?
c) Calculate the test named in b).
d) What interpretation would you give to the result calculated in c) for an alpha level of 0.05?
QUESTION 3 (10 marks)
The Judicial Commission in NSW is interested in determining the attitudes of judges and magistrates to judicial education. How large a sample would they need to determine this question at an alpha level of 0.01, to within 3%. Pilot work in Victoria suggests that 32% of judges and magistrates favour judicial education.
QUESTION 4 (20 marks)
A researcher is interested in the relationship, if any, between the length of time that a heroin user has been regularly using heroin and the number of convictions that person has. The researcher collects the following data:
Individual Length of drug use Number of convictions
1 2 5
2 8 9
3 9 12
4 5 7
5 7 4
6 3 2
7 12 13
8 10 8
9 8 10
10 6 5
a) What test would you use to determine whether there is a relationship between the two variables? Why?
b) Calculate the test named in a)
c) Would you use a one tailed test or a two tailed test? Why?
d) With an alpha level of 0.05, and your answer in c) how would you interpret this result?
PART C (15 marks)
YOU ARE REQUIRED TO ANSWER THREE (3) QUESTIONS IN THIS SECTION. EACH QUESTION IS WORTH FIVE (5) MARKS.
QUESTION 1
In examining the relationship between two interval level variables you calculate a Pearson's correlation of 0.69. What additional information do you need before you can go on to interpret what the correlation means?
QUESTION 2
What is a Type I error?
QUESTION 3
What is a two tailed test?
QUESTION 4
How do you measure the relationship between two variables which are each measured using different scales of measurement?
QUESTION 5
What is the null hypothesis and why is it important?
QUESTION 6
When would you use a Spearman's rank order correlation?
QUESTION 7
What factors are important considerations in determining the size of a random sample needed to undertake a given piece of research? Why ?
QUESTION 8
What does a negative z score indicate?
QUESTION 9
What does it mean to say that the result is not statistically significant?